Metal additive manufacturing (AM) has long struggled with reliability and quality control due in part to the prevalence of part defects. Addiguru proposes to develop and validate an intelligent system which Combines in-situ (real-time) monitoring with artificial intelligence (AI) based software to detect defects Provide feedback to Laser Powder Bed Fusion (LPBF) systems that will correct defects during the build proces This, in turn, will create a closed-loop or self-healing system. The closed-loop system can reduce defect rate by as much as 40% which is observed in Laser Powder Bed Fusion (LPBF) manufacturing process. An optical camera will capture images of each layer which will then be processed by the deep learning software to detect and characterize defects. The enhanced software will identify updated 3D printing (AM) parameters for future layers. These updated parameters in upcoming layers will be communicated back to the 3D printing (AM) machine in order to repair or significantly reduce identified defects, as well as prevent future defects. Phase I will focus conducting product-market fit as well as technical feasibility of the solution. Technical feasibility of closed-loop solution will include enhancing the capability of Addiguru's currently commercialized software, and then validating the closed-loop solution along with EWI's existing Open Architecture LPBF system. The focus will be on defect detection and correction of those defects in real time.
